Claude Sonnet 5 Launches with Introductory $2/$10 Pricing

What launched

On June 30, 2026, Anthropic released Claude Sonnet 5 at $3.00/MTok input and $15.00/MTok output. Introductory pricing of $2.00/$10.00 runs through August 31, 2026.

Sonnet 5 is Anthropic's new mid-tier model, sitting between Haiku 4-5 and Opus 4-8.

The confidence scores

Sonnet 5 ranks first or second in five of eight task types:

• -Code review: 96/100 (rank 1)
• -Code generation: 93/100 (rank 2)
• -Summarization: 91/100 (rank 1)
• -Reasoning: 90/100 (rank 2)
• -QA: 89/100 (rank 2)

For code review, Sonnet 5 scores higher than Opus 4-8 at 94/100. It costs 40% less on input.

Introductory pricing makes it the mid-tier value leader

At $2.00/$10.00 through August, Sonnet 5 undercuts every mid-tier competitor on output pricing:

• -Sonnet 5 (intro): $2.00 input, $10.00 output
• -Gemini 3.1 Pro: $2.00 input, $12.00 output
• -GPT 5.4: $2.50 input, $15.00 output
• -GPT 5.6 Terra: $2.50 input, $15.00 output
• -Claude Sonnet 4.6: $3.00 input, $15.00 output

For a code review workflow at 5,000 input and 2,000 output tokens per call:

• -Sonnet 5 (intro): (5,000/1M x $2) + (2,000/1M x $10) = $0.030/call
• -Sonnet 4-6: (5,000/1M x $3) + (2,000/1M x $15) = $0.045/call

That is a 33% savings while scoring 1 point higher on code review confidence.

After August, it still competes

At full pricing of $3.00/$15.00, Sonnet 5 matches Sonnet 4-6 on price. The upgrade path is free with better scores across every task type.

Teams on Sonnet 4-6 should switch to Sonnet 5 now while introductory pricing lasts.

See the savings for your workload at /teardown, or start tracking with a free dashboard.